ultimate-photo-digitizer/extract_photos.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
""" Author:                     Hendrik Schutter, mail@hendrikschutter.com
"""

import cv2
import sys
import os

def convert_to_avif(input, output):
    # print("input: " + str(input))
    # print("output: " + str(output))
    temp = subprocess.Popen(
        ["avifenc", "--jobs", "all", input, output], stdout=subprocess.PIPE
    )
    # get the output as a string
    # output = str(temp.communicate())
    # print(output)
    temp = subprocess.Popen(
        [
            "sync",
        ],
        stdout=subprocess.PIPE,
    )

def auto_crop_scan(src_path, output_dir):
    # read the input image
    img_src = cv2.imread(src_path)

    if not img_src:
        print("unable to read image")
        sys.exit(-1)

    print("read done")

    # convert the image to grayscale
    img_gray = cv2.cvtColor(img_src, cv2.COLOR_BGR2GRAY)
    print("convert to gray done")

    # apply thresholding on the gray image to create a binary image
    ret, thresh = cv2.threshold(img_gray, 127, 255, 0)

    # find the contours
    contours, hierarchy = cv2.findContours(
        thresh, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE
    )

    areas = list()

    for cnt in contours:
        # print(cv2.contourArea(cnt))
        areas.append(cv2.contourArea(cnt))

    areas.sort(reverse=True)

    # print(areas)

    if len(areas) >= 5:
        outer = areas[0]
        inter_min = areas[4]
        print("Outer area: " + str(outer))
        print("Inner area: " + str(inter_min))

        index = 0

        for cnt in contours:
            area = cv2.contourArea(cnt)
            if (area < outer) and (area >= inter_min):
                # compute the bounding rectangle of the contour
                x, y, w, h = cv2.boundingRect(cnt)
                # draw the bounding rectangle
                # imgView = cv2.rectangle(img,(x,y),(x+w,y+h),(0,255,0),2)

                x = x + 5  # TODO
                y = y + 5  # TODO
                w = w - 10  # TODO
                h = h - 10  # TODO

                img_crop = img_src[y : y + h, x : x + w]
                # cv2.imshow("cropped", crop_img)

                export_file_name = (
                    os.path.splitext(os.path.basename(os.path.normpath(src_path)))[0]
                    + "-"
                    + str(index)
                    + ".png"
                )
                print(export_file_name)

                export_file_path = os.path.join(output_dir, export_file_name)
                print(export_file_path)

                cv2.imwrite(export_file_path, img_crop)
                index = index + 1
                # cv2.waitKey(0)


def main():
    print("starting ...")

    destination_dir_path = -1
    source_dir_path = -1
    tmp_dir_path = -1

    for argument in sys.argv:
        if argument.startswith("destination_dir"):
            destination_dir_path = argument.split("=")[1]
        if argument.startswith("source_dir"):
            source_dir_path = argument.split("=")[1]
        if argument.startswith("tmp_dir"):
            tmp_dir_path = argument.split("=")[1]

    if (destination_dir_path == -1) or (source_dir_path == -1) or (tmp_dir_path == -1):
        print("Unable to parse config!")
        print("Example usage:")
        print(
            "     python extract_photos.py source_dir=scans/ destination_dir=photos/ tmp_dir=tmp/"
        )
        sys.exit(-1)

    for path in os.listdir(source_dir_path):
        if os.path.isfile(os.path.join(source_dir_path, path)):
            print(path)


if __name__ == "__main__":
    main()